Southern Methodist University Southern Methodist University

SMU Scholar SMU Scholar

Historical Working Papers Cox School of Business

1-1-1992

The Strategic Value of Leverage : An Exploratory Study The Strategic Value of Leverage : An Exploratory Study

Jose C. Guedes Southern Methodist University

Tim C. Opler Southern Methodist University

Follow this and additional works at: https://scholar.smu.edu/business_workingpapers

Part of the Business Commons

This document is brought to you for free and open access by the Cox School of Business at SMU Scholar. It has been accepted for inclusion in Historical Working Papers by an authorized administrator of SMU Scholar. For more information, please visit http://digitalrepository.smu.edu.

We gratefully acknowledge helpful comments received from Gerald Garvey, Richard Green, Gordon Phillips, Sheridan Titman, Ivo Welch and participants at the 1992 WFA Meetings.

THE STRATEGIC VALUE OF LEVERAGE: AN EXPLORATORY STUDY

Working Paper 92-0905*

by

Jose C. Guedes Tim C. Opler

Jose C. Guedes Department of Finance

Edwin L. Cox School of Business Southern Methodist University

Dallas, Texas 75275 (214) 692-4110

Tim C. Opler Edwin L. Cox School of Business

Southern Methodist University (214) 692-2627

* This paper represents a draft of work in progress by the authors and is being sent to you for information and review. Responsibility for the contents rests solely with the authors and may not be reproduced or distributed without their written consent. Please address all correspondence to Jose c. Guedes.

The Strategic Value of Leverage: Empirical Evidence

Abstract

This paper explores the strategic role of debt in product markets by relating cross­sectional differences in inter- and intra-industry leverage ratios to proxies for the potential for collusion and entry in 199 industries in 1989. Inter-industry regressions show little overall relation between measures of structure, contestability and leverage. Moreover, there is little evidence that the intra-industry determinants of leverage differ when the potential for strategic interaction is high. This leaves little empirical basis for broadly applicable, theoretical models where leverage policy can interact with a firm's financial condition in oligopolistic industries.

1. Introduction

A rich theoretical literature suggests that corporate capital structure choice may be

influenced by the potential for collusion and predatory behavior within an industry. For

example, Brander and Lewis (1986) show that a unilateral increase in leverage tilts

management incentives toward expansion, resulting in higher profits and an expanded

market share. In contrast, Fudenberg and Tirole (1985) and Bolton and Scharfstein (1990)

show that leverage disadvantages firms by exposing them to the risk of predation by

more liquid and less levered rivals. In reviews of this literature, Ravid (1988) and Harris

and Raviv (1991) point out that while the strategic benefits of debt in product markets

may be important in capital structure decisions, little empirical work has been carried

out in this area.

In this paper we investigate whether cross-sectional variation in firm's capital

structures is related to the potential for oligopolistic collusion and rivalrous predation.

While motivated by the theoretical literature, our analysis is exploratory in nature. We

do not test specific structural implications of existing, stylized models of the interaction

1

between product market factors and capital structure. Rather, the main question we ask

is whether measures of industry structure (i.e. concentration ratios and the number of

firms in an industry) and contestability (i.e. historical entry rates and minimum efficient

scale) can explain a significant proportion of the observed cross-sectional variation in

firm's capital structures. Our goal is to find whether significant empirical regularities

relating capital structure to strategic factors exist. After introducing a set of "traditional"

control factors which are known to be related to leverage, we test whether strategic

factors in the product market have additional power to ~xplain inter- and intra-industry

differentials in firm's use of short-term, long-:term and convertible debt. Given the

exploratory nature of our inquiry, we analyze some effects that have not been examined

by existing theories. In this way we hope to point to areas where current and future

research efforts may explain important patterns in the data.

To our surprise, the results are largely negative. We find no evidence that

differences in industry leverage ratios are related to measures of competitive conditions

within an industry. At best, strategic factors can account for only two percent of the

cross-sectional variation in inter-industry leverage ratios. Moreover, we do not find that

the traditional cross-sectional determinants of leverage across and within industries are

influenced by industry structure as is implied by several models.

Our results contrast with those of Phillips (1992a) and Spence (1985) who find

evidence of a strategic role for leverage in product markets. Phillips (1992a) finds that

increases in debt appear to have reduced price competition in three concentrated

industries that also have high entry barriers. Phillips' study differs from ours in that it

2

is dynamic in nature and uses information on prices and quantities of goods sold. He

also restricts his analysis to industries which have seen leveraged recapitalizations and

leveraged buyouts by the largest incumbents. Because we do not require price and

quantity data we are able to study a much larger group of industries. Our failure to

find a cross-sectional relation between leverage and strategic factors could be consistent

with Phillips' results if the strategic effects of leverage are transitory or limited to

industries which have seen dramatic changes in incumbent's capital structures. Spence

(1985) finds that firms in concentrated industries tend to have less leverage than other .

firms. His regressions include data on 940 firms in the 1970-74 period. It is more

difficult to account for the difference between our results and those of Spence. The

larger number of controls used in this study and differences in specifications may

explain some of the difference! In any case, Spence's results do not assign a large

empirical role to product market factors.

The remainder of the paper proceeds as follows. In Section 2 we lay out the

study design and identify empirical proxies for the potential for strategic product market

interaction. In Section 3 we provide our empirical results. In Section 4 we discuss the

implications of the results for existing models and conclude with some suggestions for

future research.

1Unreported analyses structured similar to those in Spence (1985) did not show a statistically significant, negative effect of industry concentration on leverage.

3

2. Study Design

The empirical implications of models of the strategic role of leverage in product markets

can be broadly classified by whether they involve cross-sectional differences in financial

policy across industries or within industries. Some models imply that certain types of

industries will have higher average leverage while others suggest that certain types of

firms inside an industry will have lower (or higher) leverage relative to their rivals

because of strategic factors. Thus, we estimate inter- and intra-industry leverage

regressions to separately explore these mechanisms in which strategic factors affect

firm's financial decisions. These regressions include proxies for both "traditional"

determinants of capital structure and proxies for strategic determinants of capital

structure. We use variable addition tests to determine whether regressions including

strategic proxies have greater explanatory power than regressions based on traditional

factors only. The set of traditional factors is virtually the same as that employed by

Titman and Wessels (1988).2

2.1 Inter-industry Effects

Product market-capital structure models can be classified into "debt makes you tough"

and "debt makes you weak" categories. Brander and Lewis (1986) and Maksimovic

2There are several differences in our variable set and that of Titman and Wessels (1988). First, we do not include employee quit rates because they are no longer published. Second, we do not relate firm's investment tax credits to leverage because the Tax Reform Act of 1986 repealed the Investment Tax Credit (ITC) and our sample is from 1989. Third, we use only one measure of profitability. Fourth, because the median ratio on intangibles to assets within an industry is almost always zero, we do not include this variable in inter-industry regressions.

4

(1988) are examples of the first category while Fudenberg and Tirole (1985), Bolton and

Scharfstein (1990), Phillips (1992b) and Guedes (1991) are examples of the second.

In "debt makes you tough" models, oligopolistic sellers behave aggressively,

usually in a Cournot competition game, by taking on debt.3 In a non-collusive

equilibrium as in Brander and Lewis (1986) firms in the industry raise mutually

destructive amounts of debt leading to increased output and depressed prices. In a

collusive equilibrium as in Maksimovic (1988) firms limit the amount they borrow in

order to restrain themselves from pursuing expansionary policies. The level of leverage

in oligopolistic industries thus depends on the particular type of equilibria that emerges.

For example, leverage should be lower in industries where collusion is easier to sustain.

In a cross-sectional sample of high strategic interaction industries, however, average

industry profitability and leverage should be inversely related if observations are drawn

from both equilibrium outcomes. These predictions are reversed in "debt makes you

weak" type of models such as Guedes (1991) and Phillips (1992b). Here, management

uses discretionary resources to invest in new capacity and advertising. Debt limits

available discretionary resources and thus weakens the firm's position in the product

market. Moreover, in a cross-sectional sample of industries with high potential for

strategic interaction, average industry leverage should be positively related to

profitability and negatively related to measures of investment effort.

In other models such as Fudenberg and Tirole (1985) and Bolton and Scharfstein

~his conclusion is sensitive to the type of competition assumed. John and Sundaram (1992) find the opposite result when Bertrand price competition occurs in an industry.

5

(1990) debt makes the firm weak because it makes it vulnerable to predation. These

"long-purse" models suggest that firms in industries where rivals can exercise market

power will have less debt. Also, in these models debt makes the firm weak because

asymmetric information between management and lenders leads to refunding difficulties

when profits are low. Since the risk of encountering refunding difficulties is most severe

for firms with asymmetric information problems, those firms will rationally assume less

debt in the first place. Thus, proxies for asymmetric information (e.g. R&D intensity in

the industry and size) should have a stronger negative effect on leverage in industries

where firms' possess market power.

We examine several measures of the degree of strategic interaction in an industry.

One measure is the number of firms in the industry. [See Bresnahan and Reiss (1992)].

A more traditional measure of market power is the Four-firm concentration ratio and the

Herfindahl index of industry concentration. The Department of Justice, for example,

bases much of its analysis of the anti-competitive effects of mergers on concentration

ratios. At low levels of concentration, the ability of the firm to influence its competitors'

behavior is likely to be negligible, giving leverage no strategic value.4

The sustainability of collusion is also likely to depend heavily on the con testability

of an industry by entrants [e.g. Baumol (1983)]. The Department of Justice also looks at

contestability when scrutinizing mergers. Unfortunately, it is not easy to measure the

4It would be misleading to suggest that the potential collusive behavior can readily be identified by selecting industries with concentration ratios above some critical threshold. Nonetheless, several studies suggest that collusive behavior is most prevalent in industries with Four-firm concentration ratios in excess of 50 percent. [See Geithman, Marvel and Weiss (1981)].

6

potential for entry into an industry. One factor which may raise the cost of entry is the

minimum investment needed to achieve efficient scale economies. A widely used

measure of minimum efficient scale is the average plant size within an industry [e.g.

Strickland and Weiss (1976)]. An alternative measure of contestability is the historical

rate of entry into an industry. It is possible, of course, that industries with low historical

rates of entry still experience little collusion because of pressure from potential entrants

[see Panzar and Willig (1977)]. However, it is unlikely that industries with high actual

historical entry have greater costs of entry than those with low entry rates. Thus, we use

a dummy variable for industries with low historical rates of entry as a crude proxy for

contestability.

We use these measures of industry structure and contestability to test for inter­

industry strategic effects in two ways. First, we examine their explanatory power when

used as additional independent factors over and above the Titman and Wessels (1988)

control variables. Second, we examine their influence by interacting the Titman and

Wessels controls with a dummy variable which takes the value one for high strategic

interaction industries and zero otherwise. From our previous discussion we know that

some of these interaction variables such as those based on profitability and investment

effort can be motivated by existing theoretical models. Others, however, have not been

the subject of prior theoretical work. Thus, our empirical analysis may point to some

promising areas (and also some dead ends) for future theoretical work examining the

role of financial policy in product markets.

7

2.2 Intra-industry Effects

"Debt makes you tough" and "debt makes you weak" models have opposing implications

about financial policies in industries where members enjoy market power. The former

suggest that profits, investment, and market shares are higher for firms that borrow

more extensively; the latter suggest the reverse.

As in the inter-industry case, we explore the possibility that traditional variables

have a differential impact on leverage in concentrated and unconcentrated industries.

The intra-industry regressions use deviations from the medians of both the dependent

and independent variables in order to examine the determinants of leverage differentials

within an industry. The possibility that the intra.:industry determinants of leverage

differ in concentrated industries is explored by interacting an industry concentration

dummy with the Titman and Wessels (1988) control variables.

2.3 Measures of Leverage

We examine three measures of leverage in this study: long-term debt/total assets, short­

term debt/total assets and convertible debt/total assets. Assets are measured at book

value. This is important in the context of this study since it is well known that firms

with market power tend to have higher equity values and hence lower debt/ assets at

market than other firms [Smirlock, Gilligan and Marshall (1984)).

We distinguish the determinants of long-term versus short-term debt because

short-term debt is likely to lead to more refunding difficulties in the event of poor

performance which increases firms vulnerability to predation by rivals. We also

8

conjecture that strategic factors may influence firm's use of convertible debt. For

example, the asset substitution problem which drives the Brander and Lewis (1986)

results can be substantially ameliorated with convertible debt. Thus, collusion and

aggregate industry profitability can be enhanced by the use of convertible debt. In

contrast, within concentrated industries, heavy users of convertible debt should have

lower relative profits and market shares. In addition, convertible debt may give rivals

an additional incentive to place financial pressure on a firm. For example, rivals may

lower prices in order to keep equity prices below the conversion level to maintain high

financial pressure on a firm.

2.4 Sample Selection and Data Sources

We analyze the capital structures of 3,369 U.S. firms in 1989. All of these firms had

publicly traded equity since we obtained data on firm financial characteristics from the

COMPUSTAT II PST, Full Coverage and Research files. Cases where one of the

variables required for our analysis was missing were deleted. Our data source for

information on industry characteristics is the 1989 Economic Information Systems

TRINET database. This database contains information on over 700,000 business plants

and offices (establishments) in the United States and is constructed by thousands of

phone calls each year, reviews of company documents and access to FTC business

establishment records. The coverage of the database is quite comprehensive, with

information on more than 97 percent of large public firms and on more than 80 percent

of small firms with less than 100 employees. Unlike the Industrial Census data on

9

industry concentration, this data source has information which allows to define industry

structure in all enterprise sectors of the econony. Because each record has a parent code,

an industry code, and an estimate of establishment employment the TRINET data is

ideal for measurement of variables such as market share, industry concentration and

industry entry rates.

The industry four-firm concentration ratio is defined as the fraction of employees

within a 3-digit SIC code in the largest four firms. Industries with low entry rates had

a change in number of firms of not more than 10 percent and not less than -10 percent

between 1981 and 1989. Minimum efficient scale is defined as the median establishment

size within a 3-digit SIC industry.

3. Empirical Results

3.1 Sample Description

Table 1 shows the distribution of the variables examined in this study. The firms in the

sample have a median long-term debt/ asset ratio of 20%. The median short-term

debt/ assets ratio is considerably less at 5%. Less than a fourth of the firms have any

convertible debt at all.

Table 1 shows considerable inter-industry variation in measures of the potential

for strategic product market interaction. For example, most of the 199 3-digit SIC

industries represented in our sample have low industrial concentration: the median four­

firm concentration ratio is 25%. Similarly the number of firms in most industries in the

sample is quite large. The median industry has more than 200 firms.

10

The Titman and Wessels (1988) set of control variables also exhibits high cross-

sectional variability. The interquartile range of the log of sales is 2.88 to 6.22 (or $17.8

to $502 million). Thus, most firms in our sample could best be characterized as small

to moderate. This reflects the large number of firms drawn from the COMPUSTAT Full

Coverage tape.5 Also notable is the fact that at least 25% of the firms report negative

operating earnings and a negative asset growth rate over the 1987-89 period.

While our analysis asks whether there is a statistical association between measures

of leverage and industry structure, much can be learned by examining extreme cases in

the sample. Table 2 shows the industry median of the long-term and short-term

debt/ assets ratio in the twelve 3-digit SIC industries with the greatest and the least

industry concentration in 1989. Among highly concentrated industries we see four cases

with a long-term debt/ assets ratio less than the sample median of 20% and eight cases

above this median. Most notably, there are only two industries (federal credit agencies

and cinemas) where the median long-term debt/assets ratio is more than 10 percentage

points away from the median. We also find that 9 of the 12least concentrated industries

have leverage within 10% percentage points of the sample median. Thus, it would be

fair to characterize most firms in industries with extremely high or extremely low

concentration as has having leverage positions similar to the average firm in the sample.

This simple analysis leads us to doubt that strategic factors in product markets have a

strong impact on firm's financial decisions.

~itman and Wessels (1988) did not use these firms in their study. Thus, we are studying firms which tend to be smaller than those looked at in most past studies.

11

3.2 Inter-industry Determinants of Leverage

Table 3 shows regressions predicting industry median leverage ratios using the medians

of Titman and Wessels (1988) control variables.6 There are five important results in this

table. First, given that the adjusted R2 in the regression predicting long-term debt/ assets

is 27%, the traditional controls work reasonably well in explaining cross-industry

patterns of long-term borrowing. However, these controls perform poorly in explaining

variation in short-term borrowing and convertible debt. Second, the ratio of plant &

equipment to total assets is positively associated with long-term leverage but negatively

related to short-term debt. This suggests that maturity matching of assets and liabilities

goes on at the industry level. Third, as in Titman and Wessels (1988), R&D expenses,

selling expenses, and the equipment industry dummy have an important influence on

the use of long-term debt. Fourth, cash and operating profits are negatively and

statistically significantly related to the amount of long and short-term debt. This is

consistent with the pecking order hypothesis that firms prefer internal funds to

borrowed funds. An alternative interpretation is that firms in industries facing

downsizing or adversity, which are therefore low on cash and experience op'erating

losses, tend to be more highly levered simply because they cannot increment retained

earnings. Fifth, none of the controls is found to have a significant effect on the amount

of convertible debt.

Table 4 reports regressions which add proxies for the potential for strategic

6We reduce the influence of outliers by excluding industries with median long-term debt to assets ratios more than three standard deviations from the sample median.

12

product market interaction to the previous equations. Inspection of the increment in the

adjusted R-squares of these models shows that the strategic proxies add little

explanatory power to traditional accounts of inter-industry leverage patterns. Model

specification tests reject the model with strategic variables in favor of the parsimonious

model with only the traditional controls for all the three debt types. Looking at

individual coefficients the only significant result is a positive coefficient for the low entry

dummy in the short-term debt regression. An interpretation of this finding is that,

compared to incumbents, entrants come in with less short-term debt because they face

more severe asymmetric information problems and thus are more vulnerable to

predation (since they are more exposed to refinancing difficulties in the case of poor

performance). Alternatively, this coefficient may have been statistically significant by

chance.

Measures of industry structure and contestability also contribute little to models

explaining inter-industry leverage in Table 5. This table shows regressions predicting

leverage ratios using the first six principal components from the Titman and Wessels

(1988) variables. This approach allows us to collapse the large number of controls into

a small number of orthogonalized factors and mitigate collinearity problems which are

endemic in regressions using structurally inter-related income statement and balance

sheet factors to predict leverage. The potential benefit of this approach is an increase

in the precision of the coefficient estimates for the strategic variables; the cost is a small

loss in explanatory power. The results, however, are qualitatively the same as those in

13

Table 4.1

Table 6 interacts the Titman and Wessels' controls with a dummy which takes the

value one in highly concentrated industries. These industries are defined to have four­

firm concentration ratio exceeding 50% no firm with more than 70% market share.8 As

discussed earlier, interaction variables for profitability, investment effort and proxies for

information asymmetry can be motivated by theories which discuss the role of leverage

on firm's competitive position within industries [e.g. Bolton and Scharfstein (1990),

Fudenberg and Tirole (1986), Guedes (1991), and Phillips (1992b)]. Theory suggests that

in a cross-section of concentrated industries, there should be a monotonic relation

between leverage, on one hand, and profitability and investment effort on the other.

Since this effect is over and above any impact that profitability and investment may have

on leverage in unconcentrated industries, we should find different coefficients for

profitability and investment proxies in concentrated industries.

The effect of adding the interaction variables is negligible for long-term debt and

convertible debt but significant for short-term debt (the adjusted R2 increases by 7%).

For short-term debt we find a significant positive incremental effect of industry

concentration on selling expenses/ sales and investment/ assets which may suggest a

7 Arguably, the absence of significant strategic effects is due to the large number of small firms in our sample. Perhaps the empirical importance of strategic variables is masked when there is a substantial fringe of small firms lacking market power in a concentrated industry. To address this issue we replicated Tables 4 and 5 using data on the largest four firms in each industry only. However, the qualitative character of the results did not change.

SWe impose the market share threshold to exclude monopolistic industries where debt may play less of a strategic role.

14

strategic role for short-term debt in funding expansionary investments. The lack of a

negative incremental effect on profitability, however, casts doubt on a strategic

interpretation. There is also a significant negative incremental effect for the book value

of plant, property and equipment divided by total assets which is difficult to rationalize

using existing models.9

3.3 Intra-industry Determinants of Leverage

Tables 7 and 8 present regressions which analyze the intra-industry determinants of

leverage. Table 7 contains only the Titman and Wessels' (1988) regressors while Table

8 also interacts those regressors with a dummy for high industry concentration. All

variables are measured as deviations from industry medians. As in the inter-industry

case we excluded all outliers with long-term debt/ assets ratios outside a range of plus

or minus three standard deviations around zero.

An surprising result in Table 7 is that the traditional controls do better in

accounting for intra-industry patterns of short-term borrowing than of long-term

borrowing (the adjusted-R2 are, respectively, 16% and 9%). This contrasts with the

earlier inter-industry results where the controls performed better in predicting long-term

debt (the adjusted-R2 are, respectively, 9% and 27%). In addition, pecking order

behavior appears to be stronger within than across industries. The coefficients on

~o check for robustness we replicated Table 6 for different definitions of the dummy for high concentration. Specifically, using a Four-firm concentration ratio of 37% (sample 75th percentile) and 58% (sample 95th percentile) produced no material change in the results.

15

operating profits and cash/ assets are negative and highly significant in both the long

and short-term debt regressions. Some variables which were found to be important by

Titman and Wessels are also statistically significant. This is true of the logarithm of

sales, R&D expenditures/sales, intangibles/assets and selling expenses/sales. In

addition, non-debt tax shields relative to assets is strongly negatively to leverage,

suggesting that firms substitute these shields and debt. The importance of this variable

in 1989 relative to earlier years is consistent with evidence in Givoly et. al. (1992) that

firms responded to changes in tax incentives for debt following the 1986 Tax Reform Act.

Table 8 shows that the increase in explanatory power from interacting the Titman

and Wessels' controls with a concentration dummy is quite small. The adjusted R2

increases by only 1% in the long-term debt regression and is unchanged in both the

short-term and convertible debt regressions. An F-test rejects the hypothesis that the

controls affect leverage differently in concentrated and unconcentrated industries.

On an individual basis, however, some of the controls appear to play a different

role in concentrated industries. For example, within concentrated industries, selling

expenses/ sales is strongly positively associated with long-term and convertible debt

ratios and negatively with short-term debt ratios. A "strategic" interpretation for this

finding is that firms with market power commit to an expansionary policy in the

product market by taking on more long-term and convertible debt while lowering their

short-term borrowings. Further evidence consistent with this interpretation comes from

the estimates of the incremental effect of concentration on the coefficients of operating

income/ assets and cash/ assets. The parameter estimates show that within concentrated

16

industries pecking order behavior is weaker with long-term debt but stronger with short-

term debt. This pattern may reflect the role of long-term debt in making a firm a tough

competitor in the product market. At the same time, the results for short-term debt may

reflect firm's vulnerability to predation when pressured with immediate debt payments.

If that is true, firms are more reluctant to use financial slack and operating profits to

reduce long-term debt since that undermines their competitive position in the product

market; they are also more willing to use internal funds to lower their short-term

borrowings since that reduces their vulnerability to predation by rivals.

However, there are some problems with this interpretation. First, the positive

empirical association between firm size and the amount of long-term borrowing is

weakened within unconcentrated industries. Second, the strong negative relationship

between the short-term debt ratio and size that we find within unconcentrated industries

goes away in concentrated industries. Third, the estimates of the incremental effect of

concentration on the coefficients for investment/ assets and asset growth do not indicate

that growth is associated with higher long-term and lower short-term debt ratios within

highly concentrated industries. Thus, the results are suggestive at best. On the whole,

it is difficult to argue persuasively for a strategic role for leverage based on the results

contained in Table 8.10

1'7he results in Table 8 were subjected to several sensitivity checks. We changed the concentration dummy to a Four-firm concentration ratio of 37% (sample 75th percentile) and 58% (sample 95th percentile). We measured variables by their within-industry ranks (instead of differences from their industry medians) and examined the incremental effects of concentration on rank regressions. Finally, we reestimated Table 8 using only the four largest firms in each industry. In every case we found no material change in the results.

17

4. Conclusion

There has been an upsurge of theoretical interest in the strategic role of debt in product

markets in recent years but little accompanying empirical work. We have attempted to

close the gap in the literature by undertaking an exploratory analysis of the importance

of strategic factors in firm's leverage choices. We examined the role of strategic factors

in explaining inter-industry leverage ratios and investigated whether "traditional"

determinants of leverage play a different role within highly concentrated industries as

predicted by several existing models.

On the whole, there appears to be little ground on which to argue that strategic

factors in product markets play an important role in firm's financial decisions. Proxies

for the potential for collusion have little power to explain cross-sectional variation in

leverage ratios over and above that accounted for by traditional determinants of

leverage. In the best case, we found that strategic interaction proxies increased the

explanatory power of regressions predicting inter-industry leverage by only two percent.

Similar interactions in intra-industry regressions also add little explanatory power. At

best, our results give limited evidence that debt ratios are related to the scope for

strategic interaction in the product market. The findings, even then, are difficult to

explain with existing theories of the role of debt in product markets.

While our results do not suggest a large role for models which account for capital

structure using strategic factors, it would be premature to rule such models out

altogether. For one, we have not tested specific structural implications of these models.

Nor can we ensure that the industries which are modeled have the properties of those

18

in our sample. This problem, of course, is widespread in the industrial organization

literature. Alternative approaches within the framework of the "new" empirical

industrial organization [Bresnahan and Schmalensee (1987)] call for dynamic analyses

of only a few industries at a time [e.g. Phillips (1992a)]. This limits the generalizabilty

of any conclusions drawn. Future empirical research might undertake further dynamic

analyses such as those in Phillips (1992a) in a broader range of industries. It would also

be interesting to examine firm's price and quantity responses in situtations where firm's

leverage changes endogenously as the result of recapitalizations and repurchases and in

situations where players exogenously become more or less levered because supply

shocks to an industry which affect firm's equity base.

19

Table 1. Sample Description.

Leverage ratios and leverage determinants are described by their mean, and first, second and third quartiles.

First Third Variable Mean Quartile Median Quartile

(a) Measures of leverage

Long-term debt/ assets 0.25 0.08 0.20 0.35 Short-term debt/ assets 0.10 0.01 0.05 0.13 Convertible debt/ assets 0.02 0.00 0.00 0.00

(b) Measures of industry structure (N=199)

Four-firm concentration ratio 0.29 0.18 0.25 0.36 Log of number of firms in industry 5.07 4.48 5.17 5.76 Median plant size in industry ($0000) 7.57 1.87 3.76 8.30

(c) Titman and Wessels variables (N=3914)

Selling expenses/ sales 0.34 o;2o 0.30 0.45 R&D expenses/sales 0.03 0.00 0.00 0.02 Cash & marketable securities/assets 0.10 0.014 0.047 0.13 Operating income (EBIIDA) I assets -0.032 -0.036 0.026 0.067 Depreciation/total assets 0.049 0.027 0.041 0.061 Non-debt tax shields/total assets 0.029 0.00 0.00 0.014 Log of sales 4.55 2.88 4.55 6.22 Plant, equipment & inventory I assets 0.53 0.38 0.54 0.68 Capital expenditures/ assets 0.085 0.026 0.056 0.10 Asset growth rate 1.52 -0.12 0.25 0.83 Intangibles/ assets 0.05 0.00 0.00 0.05 Standard deviation of operating income 3.31 0.48 1.43 3.25

20

Table 2. Median leverage ratios in industries with highest and lowest concentration

3-digit Long-term Short-term Four-firm Log of Median Industry SIC code debt/assets debt/assets concentration firms plant size

(a) Industries with highest concentration Communications services nee 489 0.27 0.03 0.94 1.79 9.86 Tobacco 211 0.21 0.06 0.93 1.79 62.39 Federal credit agencies 611 0.52 0.42 0.89 1.61 165.71 Greeting cards 277 0.17 0.04 0.73 3.22 4.41 Motorcycles 375 0.20 0.07 0.71 2.71 6.06 Department stores 533 0.21 0.04 0.71 4.39 1.3 Motion picture distribution 782 0.23 O.ot 0.69 2.56 21.93 Title insurers 636 0.10 0.04 0.68 3.26 4.4 Tire manufacturers 301 0.13 0.05 0.67 3.26 15.41 Cinemas 783 0.51 0.06 0.66 3.37 3.5 Photographic equipment 386 0.15 0.12 0.63 4.48 13.04 Motor vehicle manufacturers 371 0.21 0.05 0.62 6.25 23.58

(b) Industries with lowest concentration Fabricated rubber products nee 306 0.15 0.14 0.10 5.62 1.66 Home furnishing shops 571 0.19 0.07 0.09 5.43 1.23 Industrial equipment wholesale 500 0.09 0.31 0.09 6.65 2.29 Nursing homes 805 0.48 0.02 0.09 7.81 0.05 Metal work 344 0.18 0.04 0.09 6.9 2.09 Hospitals 806 0.41 0.03 0.08 8.21 3.49 Non-durables wholesale 519 0.24 0.04 0.08 6.01 5.11 Women's outerwear 233 0.22 0.06 0.08 6.44 0.74 Fabricated metal products 349 0.26 0.06 0.08 6.52 2.81 Machine tools 354 0.21 0.09 0.08 6.59 0.85 Building contractors 154 0.24 0.03 0.07 6.3 4.37 Plastic products 300 0.23 0.04 0.06 7.14 1.66

21

Table 3. Baseline inter-industry leverage regressions (industry-level)

OLS regressions predicting median debt/assets at the 3-digit SIC level for 199 industries. Regressors are industry medians of variables used in Titman and Wessels (1988).

Long term Short term Convertible debt/assets debt/assets debt/assets

Coefficient t Coefficient t Coefficient t

Intercept 0.109 2.07"' 0.155 4.43••• 0.044 0.55 Equipment industry dummy -0.038 2.1o•• 0.004 0.34 -0.011 -0.43 Selling expenses/ sales 0.122 2.01 .. 0.031 0.76 0.112 1.22 R&D expenses/sales -1.245 2.s5•• 0.335 1.03 -0.081 -0.11 Cash/ assets -0.351 2.68••• -0.241 2.75••• 0.417 1.88 Operating income/assets -0.334 1.65• -0.357 2.64••• -0.142 -0.47 Depreciation/ assets -0.078 0.18 -0.562 2.Q4•• 0.378 0.61 Non-debt tax shields/assets -1.364 2.69··· -0.075 0.22 0.853 0.85 Log of sales 0.0077 1.29 -0.0030 0.75 -0.0093 -1.07 Plant & property I assets 0.160 3.39··· -0.082 2.60••• 0.106 1.57 Investment/ assets 0.033 0.16 0.086 0.62 -0.116 -0.38 Asset growth 0.019 0.85 -0.0029 0.19 0.025 0.58 Standard deviation of income 8.3E-4 0.19 -0.0025 0.85 -0.009 0.95

Adjusted R2 0.27 0.09 0.02

••• Coefficient is significant at the 1% level. •• Coefficient is significant at the 5% level. • Coefficient is significant at the 10% level.

22

Table 4. Inter-industry regressions with strategic factors

OLS regressions predicting median debt/assets at the 3-digit SIC level for 199 industries. Regressors are industry medians.

Long term Short term debt/assets debt/assets

Coefficient t Coefficient t

(a) Variables used in Titman and Wessels (1988)

Intercept 0.033 0.37 0.204 3.49••• Equipment industry dummy -0.037 2.03 .... 0.0040 0.33 Selling expenses/sales 0.136 2.14 .... 0.035 0.83 R&D expenses/sales -1.368 2.74 ...... -0.213 0.64 Cash/ assets -0.342 2.6Q!t•• -0.245 2.81··· Operating income/ assets -0.329 1.62• -0.366 2.72••• Depreciation/ assets -0.096 0.23 -0.520 1.86• Non-debt tax shields/assets -1.277 2.46 .... -0.080 0.23 Log of sales 0.0088 1.36 -0.0022 0.52 Plant & property I assets 0.158 3.31··· -0.084 2.67!t• Investment/ assets 0.074 0.35 0.082 0.58 Asset growth 0.016 0.70 -2.3E-4 0.01 Standard deviation of income 0.0022 0.48 -0.0034 1.15

(b) Variables which identify the potential for strategic product market interaction

Low entry dummy Four-firm concentration ratio Number of firms in industry Minimum efficient scale

F-test of the joint significance of product market interaction variables

Adjusted R2

Increment in R2 over the baseline model (Table 3)

-0.011 0.79 0.034 0.48 0.011 1.06 -6.6E-5 0.09

0.58

0.26

-1%

..,.,. Coefficient is significant at the 1% level. " Coefficient is significant at the 5% level. • Coefficient is significant at the 10% level.

23

0.022 2.37!t• -0.065 1.39 -0.0085 1.19 9.0E-6 0.01

1.99•

0.11

+2%

Convertible debt/assets

Coefficient t

0.156 1.10 -0.0087 0.32 0.088 0.89 0.030 0.04 0.357 1.55

-0.139 0.45 0.543 0.85 0.872 0.85

-0.0077 0.81 0.098 1.43

-0.138 0.44 0.028 0.64

-0.0091 0.85

-0.024 1.18 -0.076 0.75 -0.015 0.93 -4.5E-4 0.47

0.55

0.01

-1%

Table S. Inter-industry regressions using orthogonalized traditional factors and strategic variables

OLS regressions predicting median debt/assets at the 3-digit SIC level for 199 industries. The first set of regressors industry medians are the first six principal components extracted from the Titman and Wessels (1988) variables. The second set of regressors consists of proxies for product market interaction potentiaL

Long-term debt/assets

Coefficient t Coefficient t

(a) Factors based on mriables used in Titmlln & Wessels (1988)

Intercept 0.225 34.1 .... 0.191 2.62•• Factor 1 -0.013 1.92• -0.013 1.87* Factor 2 0.036 5.42••• 0.036 5.03••• Factor 3 -0.029 4.36··· -0.029 4.v-·· Factor 4 -0.020 3.08••• -0.022 3.15••• Factor 5 -0.017 2.57" -0.016 2.30-• Factor 6 -0.0012 0.17 -0.0014 0.20

(b) Variables which identify the potential for product mllrket inreraction

Low entry dummy Four-firm concentration ratio Number of firms in industry Minimum efficient scale

F-test of the joint significance of product market interaction variables

Adjusted R2 0.24

••• Coefficient is significant at the 1% level. •• Coefficient is significant at the 5% level. • Coefficient is significant at the 10% level.

-0.014 0.032 0.0061 -2.2E-4

1.01 0.45 0.56 0.31

0.,46

0.23

Short-term debt/assets

Coefficient t

0.061 13.,.... 0.0069 1.57

-0.0081 1.84• -0.0073 1.65• -0.0055 1.25 -0.010 2.27" -0.0020 0.47

0.05

Coefficient t

0.122 0.0065

-0.0082 -0.0068 -0.0033 -0.012 -0.0012

0.021 -0.060 -0.010 -8.8E-6

24

2.52•• 1.45 1.74• 1.53 0.73 2.ss••• 0.28

2.20-• 1.30 1.41 0.01

1.82

0.06

Convertible debt/assets

Coefficient t Coefficient t

0.109 n.3••• 0.235 2.23" 0.021 2.03•• 0.018 1.64• 0.0087 0.93 0.011 1.07 0.021 1.77- 0.019 1.56 5.0E-4 0.05 0.0029 0.28 7.8E-4 0.05 0.0026 0.18

-1.8E-4 0.01 0.0021 0.20

-0.030 1.48 -0.090 0.91 -0.016 1.03 -7 .3E-4 0.78

0.86

0.03 0.02

Table 6. Inter-industry regressions with incremental coefficients for concentrated industries

OLS regressions predicting median debt/assets at the 3-digit SIC level for 199 industries. Regressors are industry medians. The incremental estimates for concentrated industries are estimated as the variable shown multiplied by a dummy variable which denotes industries with a four-firm concentration ratio in excess of 50% where no firm has a market share exceeding 70%.

Long tenn Short tenn Convertible debt/assets debt/assets debt/assets

Coefficient t Coefficient t Coefficient t

(a) Baseline Estimates

Intercept 0.087 1.54 0.150 4.15••• 0.026 0.30 Equipment industries dummy -0.031 t.66• 0.0055 0.45 -0.002 0.09 Selling expenses/ sales 0.180 2.691t>tlt -0.013 0.31 0.091 0.87 R&D expenses/sales -1.514 2.95 ... -0.203 0.62 -0.113 0.15 Cash/ assets -0.466 3.25••• -0.165 1.78• 0.440 1.76• Operating income/ assets -0.565 2.26•• -0.304 1.88• -0.337 0.89 Depreciation/ assets 0.420 0.69 -0.979 2.51•• 0.782 0.79 Non-debt tax shields/assets -1.024 1.38 -0.095 0.20 0.172 0.09 Log of sales 0.0094 1.30 -0.0035 0.76 -0.002 0.21 Plant & property I assets 0.099 1.73• 0.012 0.33 0.056 0.67 Investment/ assets 0.258 0.77 -0.434 2.03 .. 0.086 0.14 Asset growth 0.023 0.71 0.018 0.85 -0.001 0.03 Standard deviation of income 0.0072 0.96 -0.0001 0.04 -0.008 0.61

(b) Incremental estimates for concentrated industries

Selling expenses/ sales -0.218 t.83• 0.130 1.7Qit 0.170 0.98 R&D expenses/sales 0.169 0.10 -0.073 0.07 -0.898 0.41 Cash/ assets 0.487 1.46 0.015 0.07 -0.018 0.03 Operating income/assets 0.495 1.04 0.056 0.18 0.599 0.78 Depreciation/ assets -0.280 0.31 0.242 0.42 -0.742 0.54 Non-debt tax shields/assets -0.532 0.51 -0.055 0.08 0.772 0.33 Log of sales -4.5E-4 0.04 0.0039 0.60 -0.021 1.40 Plant & property I assets 0.118 1.28 -0.187 3.14 ... 0.133 0.99 Investment/ assets -0.240 0.56 0.782 2.84•• -0.235 0.34 Asset growth -0.023 0.47 -0.031 1.00 0.055 0.50 Standard deviation of income -0.010 1.02 -0.0026 0.43 -5.6E-5 0.01

Adjusted R2 0.28 0.16 -0.01 Increment in R2 over baseline +1% +7% -3%

... Coefficient is significant at the 1% level. •• Coefficient is significant at the 5% level. • Coefficient is significant at the 10% level.

25

Table 7. Baseline intra-industry leverage regressions (firm-level)

OLS regressions predicting intra-industry leverage ratios with assets measured at book for 3,369 firms in 1989. All variables represent deviations from 3-digit SIC industry medians. The regressors are those used in Titman and Wessels (1988).

Long term Short term Convertible debt/assets debt/assets debt/assets

Coefficient t Coefficient t Coefficient t

Intercept 0.043 9.4o--•• 0.049 15.48 ...... 0.019 2.45 .... Dummy for equipment industries 0.012 1.64 .. 0.0038 0.77 0.015 1.41 Selling expenses/ sales 0.029 1.99 .... -o.020 1.97"" .. 0.015 0.53 R&D expenses/sales -o.086 1.31 -o.149 3.3<r" .... -o.097 0.96 Cash/assets -o.204 5.93 ...... -o.236 9.94 ...... 0.221 4.55 ...... Operating income/ assets -o.224 7 .3<r" .... -o.344 16.26 ...... -o.084 1.93 Depreciation/ assets 0.029 0.26 -o.144 1.9<r" 0.371 2.49* .. Non-debt tax shields/assets ..0.322 6.72 ...... 0.049 1.47 -o.132 1.12 Log of sales 0.0088 4.77"""' .. -o.0072 5.62 ...... -o.013 4.82"' .... Plant & property I assets 0.0041 0.17 6.4E-4 0.03 0.0064 0.18 Investment/ assets 0.022 0.50 -o.104 3.42 ...... -o.093 1.14 Asset growth 0.0011 1.58 3.4E-4 0.67 -7.9E-4 0.79 Intangibles/ assets 0.472 4.18 .... -o.014 0.17 -o.041 0.50 Standard deviation of income 0.0001 0.76 -1.3E-5 0.13 -3.8E-4 1.96,.,.

Adjusted R2 0.09 0.16 0.11

,...,.. Coefficient is significant at the 1% level. ,.,. Coefficient is significant at the 5% level. ,. Coefficient is significant at the 10% level.

26

Table 8. Intra-industry regressions with incremental coefficients for concentrated industries

OLS regressions predicting intra-industry leverage ratios with assets measured at book value for 3,336 firms in 1989. All variables represent deviations from 3-digit SIC industry medians. The incremental estimates for concentrated industries are estimated as the variable shown multiplied by a dummy variable which denotes industries with a four-firm concentration ratio in excess of 50% where no firm has a market share exceeding 70%.

Long term Short term Convertible debt/assets debt/assets debt/assets

Coefficient t Coefficient t Coefficient t

Intercept 0.042 9.09 ...... 0.049 15.49••• 0.017 2.19 .... Dummy for equipment industries 0.014 1.94 .. 0.0035 0.70 0.013 1.23 Selling expenses/ sales 0.0064 0.41 -o.0073 0.68 -o.028 0.85 R&D expenses/sales -o.034 0.51 -o.153 3.34••• -o.059 0.57 Cash/ assets -o.230 6.2011-.... -o.235 9.1711-.... 0.250 4.78••• Operating income/assets -o.233 7.19··· -o.320 14.22•** -o.097 2.11•• Depreciation/ assets 0.179 1.34 -o.128 1.38 0.546 3.Q6••• Non-debt tax shields/assets -o.318 6.33··· 0.047 1.34 -o.094 0.65 Log of sales 0.011 5.44••• -o.0089 6.1711-.... -o.012 4.03··· Plant & property I assets -o.016 0.58 0.013 0.67 -o.0096 0.24 Investment/ assets 0.0035 0.07 -o.124 3.54* .... -o.059 0.64 Asset growth 0.0009 1.24 3.7E-4 0.69 -o.0010 0.99 Intangibles/ assets 0.555 4.49*•* -o.076 0.88 -o.059 0.66

(b) Incremental estimates for concentrated industries.

Standard deviation of income 7.3E-5 0.48 -2.3E-5 0.22 -4.6E-4 2.25•• Selling expenses/ sales 0.192 4.21•** -o.099 3.13 ...... 0.135 1.96 .... R&D expenses/sales -1.866 3.96··· 0.043 0.13 0.850 0.96 Cash/ assets 0.240 2.53 .... -o.016 0.24 -o.135 0.99 Operating income/ assets 0.093 1.04 -o.193 3.13*** -o.070 0.48 Depreciation/ assets -o.S24 2.18*• -o.100 0.60 -o.687 1.94* Non-debt tax shields/assets -o.104 0.73 0.058 0.58 0.229 0.84 Log of sales -o.oos5 1.24 0.0071 2.31** -o.0065 0.88 Plant & property I assets 0.086 1.43 -o.042 1.01 0.124 1.22 Investment/ assets 0.087 0.86 0.083 1.19 -o.092 0.43 Asset growth 0.0030 1.16 -3.4E-4 0.19 0.0042 0.48 Intangibles/ assets -o.s63 1.91• 0.354 1.73• 0.169 0.76 Standard deviation of income 0.0003 0.71 6.9E-5 0.23 3.2E-4 0.50

Adjusted R2 0.10 0.16 0.11 Increment in R2 over baseline model (Table 7) +1% 0% 0%

...... Coefficient is significant at the 1% level. •• Coefficient is significant at the 5% level. • Coefficient is significant at the 10% level.

27

REFERENCES

Baumol, William J., 1983, Contestable markets: An uprising in the theory of industry structure, American Economic Review 72, 1-15.

Bolton, Patrick and David S:Scharfstein, 1990, A theory of predation based on agency problems in financial contracting, American Economic Review 80, 93-106.

Brander, James A. and Tracy R. Lewis, 1986, Oligopoly and financial structure, American Economic Review 76, 956-970.

Bresnahan, Timothy and Peter Reiss, 1991, Entry and competition in concentrated markets, Journal of Political Economy 99, 977-1009.

Bresnahan, Thnothy and Richard Schmalensee, 1987, The empirical renaissance in industrial organization: An overview, Journal of Industrial Economics 35, 371-378.

Fudenberg, Drew and Jean Tirole, 1986, A 'signal-jamming' theory of predation, Rand Journal of Economics 17, 366-376.

Galai, Dan and Ronald Masulis, 1976, The option pricing model and the risk factor of stock, Journal of Financial Economics 3, 53-81.

Geithman, Frederick E., Howard P. Marvel and Leonard W. Weiss, 1981, Concentration, price and critical concentration ratios, Review of Economics and Statistics 63, 346-353.

Givoly, Dan, Carla Hayn, Aharon R. Ofer and Oded Sarig, 1992, Taxes and capital structure: Evidence from firms' response to the Tax Reform Act of 1986, Review of Financial Studies 5, 331-355.

Guedes, Jose, 1991, Free cash flow in oligopolies, Working Paper, E. L. Cox School of Business, Southern Methodist University, Dallas, TX.

Harris, Milton and Arthur Raviv, 1991, Theories of Capital Structure, Journal of Finance 46, 297-355.

Maksimovic, Vojislav, 1988, Capital structure in repeated oligopolies, Rand Journal of Economics 19, 389-407.

Panzar, John C. and Robert D. Willig, 1977, Free entry and the sustainability of natural monopoly, Bell Journal of Economics 8, 1-22.

28

Phillips, Gordon M., 1992a, Increased debt and product market competition: An empirical analysis, Manuscript, Krannert School of Management, Purdue University.

Phillips, Gordon M., 1992b, Financing investment and product market competition, Manuscript, Krannert School of Management, Purdue University.

Ravid, S. Abraham, 1988, On interactions of production and financial decisions, Financial Management 17, 87-99.

Smirlock, Michael, Thomas Gilligan and Wiliam Marshall, 1984, Tobin's q and the structure-performance relationship, American Economic Review 74, 1051-1060.

Spence, A. Michael, 1985, "Capital structure and the corporation's product market environment," in B. Friedman (ed.), Corporate Capital Structures in the United States, University of Chicago Press, Chicago, IL.

Sundaram, Anant K. and Kose John, 1992, Product market games and signalling models in finance: Do we know what we know?, Working paper, Amos Tuck School, Dartmouth College.

Titman, Sheridan and Roberto Wessels, 1988, The determinants of capital structure choice, Journal of Finance 43, 1-19.

29

Note: The £ollowing is a partial list of papers that are currently available in the Edwin L. Cox School of Business Working Paper Series. When requesting a paper, please include the Working Paper number as well as the title and author(s), and enclose pay­ment of $2.50 per copy made payable to SMU. A complete list is available upon request from:

Business Information Center Edwin L. Cox School of Business

Southern Methodist University Dallas, Texas 75275

90-0101

90-0201

90-0701

90-1001

90-1002

90-1003

90-1201

91-0101

91-0701

91-0901

91-0902

91-1001

91-1002

91-1201

"Organizational Subcultures in a Soft Bureaucracy: Resistance Behind the Myth and Facade of an Official Culture," by John M. Jermier, John W. Slocum, Jr., Louis w. Fry, and Jeannie Gaines

"Global Strategy and Reward Systems: The Key Roles of Management Development and Corporate Culture," by David Lei, John W. Slocum, Jr., and Robert w. Slater

"Multiple Niche Competition - The Strategic Use of CIM Technology," by David Lei and Joel D. Goldhar

"Global Strategic Alliances," by David Lei and John W. Slocum, Jr.

"A Theoretical Model of Household Coupon Usage Behavior And Empirical Test," by Ambuj Jain and Arun K. Jain

"Household's Coupon Usage Behavior: Influence of In-Store Search," by Arun K. Jain and Ambuj Jain

"Organization Designs for Global Strategic Alli­ances," by John W. Slocum, Jr. and David Lei

"Option-like Properties of Organizational Claims: Tracing the Process of Multinational Exploration," by Dileep Hurry

"A Review of the Use and Effects of Comparative Advertising," by Thomas E. Barry

"Global Expansion and the Acquisition Option: The Process of Japanese Takeover Strategy in the United States," by Dileep Hurry

"Designing Global Strategic Alliances: Inte­gration of Cultural and Economic Factors," by John w. Slocum, Jr. and David Lei

"The Components of the Change in Reserve Value: New Evidence on SFAS No. 69," by Mimi L. Alciatore

"Asset Returns, Volatility and the Output Side," by G. Sharathchandra

"Pursuing Product Modifications and New Products: The Role of Organizational Control Mechanisms in Implementing Innovational Strategies in the Pharmaceutical Industry," by Laura B. Cardinal

92-0101

92-0301

92-0302

92-0303

92-0304

92-0305

92-0306

92-0401

92-0402

92-0501

92-0502

92-0503

92-0601

92-0701

92-0702

"Management Practices in Learning Organizations," by Michael McGill, John W. Slocum, Jr., and David Lei

"The Determinants of LBO Activity: Free Cash Flow Vs. Financial Distress Costs," by Tim Opler

"A Model of Supplier Responses to Just-In-Time Delivery Requirements," by John R. Grout and David P. Christy

"An Inventory Model of Incentives for On-Time Delivery in Just-In-Time Purchasing Contracts," by John R. Grout and David P. Christy

"The Effect of Early Resolution of Uncertainty on Asset Prices: A Dichotomy into Market and Non­Market Information," by G. Sharathchandra and Rex Thompson

"Conditional Tests of a Signalling Hypothesis: The Case of Fixed Versus Adjustable Rate Debt," by Jose Guedes and Rex Thompson

"Tax-Loss-Selling and Closed-End Stock Funds," by John W. Peavy III

"Hostile Takeovers and Intangible Resources: An Empirical Investigation," by Tim C. Opler

"Morality and Models," by Richard 0. Mason

"Global Outsourcing of Information Processing Services," by Uday M. Apte and Richard 0. Mason

"Improving Claims Operations: A Model-Based Approach," by Uday M. Apte, Richard A. Cavaliere, and G. G. Hegde

"Corporate Restructuring and The Consolidation of u.s. Industry," by Julia Liebeskind, Timothy C. Opler, and Donald E. Hatfield

"CatalogForecasting System: A Graphics-Based Decision Support System," by David V. Evans and Uday M. Apte

"Interest Rate Swaps: A Bargaining Game Solution," by Uday Apte and Prafulla G. Nabar

"The Causes of Corporate Refocusing," by Julia Liebeskind and Tim C. Opler

92-0801

92-0901

92-0902

92-0903

92-0904

"Job Performance and Attitudes of Disengagement Stage Salespeople Who Are About to Retire," by William L. Cron, Ellen F. Jackofsky, and John w. Slocum, Jr.

"Global Strategy, Alliances and Initiative," by David Lei and John W. Slocum, Jr.

"What's Wrong with the Treadway Commission Report? Experimental Analyses of the Effects of Personal Values and Codes of Conduct on Fraudulent Financial Reporting," by Arthur P. Brief, Janet M. Dukerich, Paul R. Brown and Joan F. Brett

"Testing Whether Predatory Commitments are Credible," by John R. Lott, Jr. and Tim C. Opler

"Dow Corning and the Silicone Implant Contro­versy," by Zarina S. F~ Lam and Dileep Hurry